saw the SP15 groupbuy thread and my initial excitment tapered out a bit when i realised what the idea behind it was.
I thought it was an SP10 fit to the case of an SP30, thereby giving more room for airflow, allowing larger (ie: lower rpm/noise) fans, and maintain a reasonable power demand for a full rack of them (50kW/server rack gets some crazy looks from the colo people).
unfortunately, its just half of an SP30 in an SP30 case.
perhaps fitting SP10 hardware to a larger case, such as putting 3TH in a 3kW 2U or even 3U, would eliminate the noise and density concerns (mostly noise concerns, which could be drastically reduced using 60-80mm fans instead of 40mm)
|
|
|
I will soon transfer to cloud mining after knccloud enable.
you're gonna send these assholes even more money? www.bitguy.comstart hashing today ^ Obviously I miss something, but: how can you ever get a positive ROI with these mining contracts? You don't... ROI is for the people selling the contracts. too true. every mining ontract is designed so that the seller will pay off the hardware costs, power/rent, and make a small margin at a minimum. so while you take the gamble of whether it will be profitable, the contract seller IS profitable
|
|
|
I am curious...EVGA has 8 PCIe cables. Are they split cables or did you use split extenders to run 5 (10 connectors) underclocked? What is the box? Is it some kind of datacenter box or something simpler? I am intrigued. The evga came with two split and two single i "borrowed" two split from another psu. so yes i used split cables. edit : the box is a corsair. all undervolted another simple option if you dont have PCIE splitters is to plug the singular PCIe into the primary hashing board, then run a pair of 14awg wires between the screw terminals of the primary and secondary hashing boards. (a pair of 16awg may do the trick if undervolted)
|
|
|
they could put the resistors in and sell the modded ones as s-1A's
low power model.
better yet they could put in 8 pots 3k ohm to 4.47k ohm and sell them as adjustable models
ones with pots would be great - I would like to see them release a more compact 'S1A' with slightly smaller heatsinks closer together and an initial 150GH/200W spec. This could drop the costs slightly and handle the first small undervolting step for the consumer
|
|
|
ive got a feeling the other 20TH that was expected to be online now might just be in the same boat as the 737GH of avolon gear he had since last year - silently mining away in an account without proper tracking or dividends. of course, that could be total FUD - but we have nothing to believe otherwise. A basic audit of the DC (# of racks, # of boards per rack, costs, profits) would clear things up but thats clearly not happening
|
|
|
Perfect analogy. It is insanely loud. Just like the data centers. how loud is LOUD?
are we talking the volume of an S2? 4-6 S1 units? or is this vaccuum-cleaner loud?
Louder. You'd probably need 40 S2s to match the noise pressure of a single SP10. christ. Im planning on installing it in its own location where noise isnt an issue, but if the answer had been less of a resounding LOUD I may have been able to dd it to my current collection of 16 S1 and a S2 unit (which are approaching the limits of noise i can handle)
|
|
|
It creates a coil like that, and inductive currents can blow both ends. I mean the miner controller + the rooter.
Wow, rly ? U think there is enough inductive current to blow something ? What u think how long cable needs to be ? And do u have some actual proof that did happen to somebody ? Only asking coz i do internet networks setup, left cables like that if they are to long many times . Dont worry about it. While in theory current is induced, it is soooo low u cant really measure it. You get like 12V / 1A from induction when hanging a huge coil below a 10000V landline. So yeah, the odds that a meteorite would crash into your home is probably higher than you blowing anything out because of induced electrical currents. Yeah that is what i thought so . I know in theory it can happen but, i mean that like 10m of UTP cable cant do no harm. You have no idea how a thunder can screw you by inductive fields. Happened to my shop last year, all cameras, printers and fax, usb sides on pcs got blown by a thunder hit a mile away. I dont care if you dont do your research, it is your equipment do as you wish. ps. Micro, you are an expert in network setups, huh? Do you have some actual proof? I can only assume you mean lightning. thunder is a bunch of sound waves with virtually zero effect on electronics. lightning should have no effect unless it somehow messed with your grounding - a coiled lan cable will not cause issues. its pretty common to coil cables if they are longer than necessary. that said, some cable types and data types (looking at you, analog) can suffer a little bit of crossfeed if there are too many wires in an area. However, this shouldnt be an issue for a Cat5+ cable
|
|
|
IMHO go with whatever you find cheapest. My two suggestions:
1) L6-30 outlets and PDU. 30A gives you about 5kW of safely usable power. i suggest buying the outlets ($50 new) on ebay for <$20.
2) use regular 120V components but swap a neutral for a hot. Its the absolute cheapest and easiest way and is safe for hardware (not sure if its safe for a surge protector?). However, this is very much against code. if someone plugged in a 120V appliance not knowing it was a 208V circuit they would explode. or whatever happens when a basic 120V appliance is extremely overloaded in a split second...
|
|
|
how loud is LOUD?
are we talking the volume of an S2? 4-6 S1 units? or is this vaccuum-cleaner loud?
|
|
|
i upgraded to a 2000W and overclocked to 1100GH
What frequency setting for 1100 ? Im running 215.625MHz (any higher was >1% errors) and the S2 says 1100GH average but pool says 1068GH average. doing the math now, its current error rate is 1.06% so i may need to tune it down to 212.5MHz
|
|
|
Along with everyone else, I'm hoping he delivers a dividend for gen 3. But do I expect that if asicminer really succeeds and starts to generate significant continuous revenue and expand into various other markets that the current shareholder arrangement will continue? Not a chance in hell.
What makes you say that? Did General Electric (GE) change its structure after turning 100yr old and becoming multibillion dollar company? Did Microsoft or Google or Apple screw over their shareholders when they made it big? Microsoft technically became huge by shafting IBM. (some would say outsmarting)
|
|
|
Are you up to Fire Code on that building? Since it is connected to your home.
Code specifications would be great if given. And what code specs did you go by or did you overkill on the codes? Overkill on codes is a great thing.
Thats a good question.... probably not up to code exactly. The shed sits on an external slab backed up to the exterior of the house which is cement board (deem as a non-combustible material). We also have a nest protect in there for fire/smoke warnings. In fact it went off last night I think because there was a fire somewhere outside (you could smell it) and the fresh air intake brought some in. Just a warning, google recalled a lot of nest. Becareful, man. Hope you live by yourself and not with children, wife & pets or putting someones life at risk including yours. http://www.informationweek.com/cloud/software-as-a-service/googles-nest-recalls-smoke-detector/d/d-id/1269110Only recalled because of the wave to dismiss option. ^ha, how did that feature slip into the final product? *nest senses fire - send alert* *humans rush into room on fire and flailing - wave to dismiss activated - returning to standby*
|
|
|
I was noticing that the hashrate for my antminers had been dropping off. Logged into them and noticed they had some pretty high error rates: 18 Hours 30 Minutes Ant1 275 = 12.763361189114817696879475815092 Ant2 250 = 0.10015846041845231355797176812215 Ant3 325 = 4.4344988816100151774557453763173 Ant4 275 = 8.9107904850450710092977864238739 Pulled them onto my test bench to remeasure resistance values to see if it had changed any. Sure enough all of the miners, but one had massive resistance drops. Used a pencil to readjust them back to nominal and things are running properly again. Ant1 2.675 kOhms .802v top / .802v bottom 2.480 kOhms .800v top / .800v bottom 2.270 kOhms 2.674 kOhms - R3 Resistance 2.838 kOhms 0.851v top / 0.847v bottom R66 Resistance 2.842 kOhms 0.858v top / 0.853v bottom R38 Resistance 2.944 kOhms 0.861v top / 0.856v bottom = *added* R52 Resistance 2.876 kOhms 0.859v top / 0.854v bottom
2.705 kOhms 2.566 kOhms 2.633 kOhms 2.730 kOhms - R3 Resistance 2.860 kOhms 0.854v top / 0.850v bottom R66 Resistance 3.109 kOhms 0.881v top / 0.876v bottom = *added* R38 Resistance 2.964 kOhms 0.867v top / 0.863v bottom = *added* R52 Resistance 2.850 kOhms 0.853v top / 0.849v bottom
Frequency = 275 Power Consumption = 176 watts Roughly = 139.73 GH 30 Minute Error Rate = 0.55135273753045262213104244133863 Ant2 Ant3 3.640 3.578 3.620 3.259 R3 Resistance 3.870 1.010v top / 1.004v bottom R66 Resistance 4.000 1.035v top / 1.028v bottom R38 Resistance 3.900 1.014v top / 1.008v bottom R52 Resistance 3.930 1.010v top / 1.006v bottom = Added - 3.478 3.475 3.750 3.560 R3 Resistance 3.960 1.008v top / 1.001v bottom = Added R66 Resistance 3.920 1.008v top / 1.001v bottom = Added R38 Resistance 3.920 1.014v top / 1.007v bottom R52 Resistance 3.999 1.012v top / 1.007v bottom = Added
Frequency = 350 Power Consumption = 280 watts Roughly = 167.65 GH 30 Minute Error Rate = 0.2453540629175270253856431434471 Ant4 2.670 2.599 2.533 2.481 R3 Resistance 2.916 kOhms .860v top / .856v bottom = Added R66 Resistance 2.875 kOhms .860v top / .856v bottom = Added R38 Resistance 2.855 kOhms .851v top / .847v bottom = Added R52 Resistance 2.959 kOhms .864v top / .860v bottom = Added - 2.659 2.912 2.591 2.674 R3 Resistance 2.871 kOhms .854v top / .848v bottom = Added R66 Resistance 2.912 kOhms .865v top / .860v bottom R38 Resistance 2.875 kOhms .861v top / .855v bottom R52 Resistance 2.989 kOhms .870v top / .865v bottom
Frequency = 275 Power Consumption = 174 watts Roughly = 139.49 GH 1 Hour Error Rate = 0.8297014898156421524827221502945 pencil mods have a tendency to 'bake in'. Within the first 24 hours the resistance may drop slightly and thus the voltage may slip below what worked at first
|
|
|
so after sealing up the wall we're sitting at 86 degrees f on the cool side and 124 degrees f on the hot side. This is with the 32 miners running, and it isn't extremely hot outside yet. Shaxs and I are going to tinker with individual shields for each miner to try and contain the heat they put off so it's funneled directly to the hot side instead of outwards towards the other miners.
This is like the never-ending project...
sounds pretty fun and successful though! 32 miners is a lot of heat - are they overclocked or undervolted? whats the outside temperature? and as mentioned, what is the noise level like outside the shed - 32 S1 units make a LOT of noise when you dont have a wall between you and them
|
|
|
there is documentation in the github manuals for the U1 usb device. It relies on a weird mathematical method of several combined binary strings from the hex code. Once sorted out its *relatively* straight forward to calculat eany frequency you want, to about 1.75MHz precision
Thanks for the information. Fancy math indeed! I wasn't able to "reverse engineer" the calculation (ie take a desired frequency and find the Hex value that I should use on the mining script) but I was able to calculate valid values for all the 4 digit Hex numbers. I found a lot that were supposed to be valid but they didn't work when using my U2 with BFGMiner. Maybe my math was wrong or maybe the U2 don't take all valid values in the formula. It's more probable my math was wrong, since I got some weird "valied" values, like 306.1429 or something like that. Will keep experimenting and create a list of valid HEX/frequency combinations. FM i threw out the paper i was working on, but the biggest factor is to determine your desired frequency and work backwards. Final Freq (lets say its 300) = 25Mhz (this wont change) x [N/M] (M being 1, 2, or 4 based on the last hex value being a 0,1,or 2 respectively). That makes calculating 'N' from the formula a bit easier. Once you do the method 2 or 3 times its easier to predict the hex code you will need. (for example, the 200-225 range is generally a x0yz with x being 4 or 5 in most of my work for that frequency range. and using the U1 or U2 as a testbench is a great way to see if it will run and check if the average hashrate is what you expected.
|
|
|
has anyone attempted upgrading to a new 1300 Watt PSU?
i upgraded to a 2000W and overclocked to 1100GH
|
|
|
So your power cost including delivery is around 10ct? is that correct?
also, cooling typically adds around 30% more power consumption when running at full load, so it isnīt that much more power cost.
It is a necessity in most parts of the US anyway.
From my experience, antminers arenīt that sensitive to a higher intake temperature of say 30°C.
guess that's what i meant to ask above. OP: would you skip the AC unit in favor of simply using outside 30C air (at peak hours) in your shed, and if not (due to limits of the current ventilation CFM?) do you think a shed with larger vents and fans would be capable? second part: is running an AC on a very small amount of intake air very effective if the air is being used and vented out hot within a minute or less? and does doing it that way mean it will add less than 30% (the general figure for closed-circuit cooling of heat) because the intake air (30C) is still significantly cooler than the exhaust (45C)? lastly: whats the noise like? does it sound like a pool shed or does it sound like a vaccuum/box of fans when you stand outside or nearby? (ie: would it annoy the neighbours or be obvious)
|
|
|
Has anyone done this to get 200GH per ant 100GH per blade etc with reduced voltage. So less voltage more GH am thinking to do but just checking here first to see if anyone's managed to reduce the voltage and have a good hash speed
Imo the 150GH/210W target is pretty ideal. Good payoff between time spent modding and the power savings. If you did 180GH/300W you would find yourself re-modding a few weeks later. (that said, 3.9-4.0 kohm might give 180GH / 300W if you are lucky)
|
|
|
http://eligius.st/~wizkid057/newstats/userstats.php/1DJpsvnM7xTnQbWEhLYyCyfxQyxwupEzCathe story i see: 170TH was originally promised by the end of may that was changed to 130-150TH after fixing the 3rd AC on May 13th, at which time there was 70TH the DC still has yet to break 100TH, and if anything has gone from running ~95TH for a few days back down to ~88TH. this is not growing as expected, and a detailed update with photos would be nice. Its too easy to assume that a malignant DC operator could 'suffer outtages and/or hashrate loss' when in fact diverting hashrate to private pools temporarily. a 99% uptime should be expected, and if not, explained
|
|
|
Hello,
I have a new Antminer S1 and when I turn it on after 5 minutes the whole room smells like burning plastic. My other Antminer does not smell like that, and never did.
Does anyone have the same issue? Has it gone away? Any ideas?
It looks fine and runs cool. Better than my older one, but a terrible smell.
It will go away. I believe it's the smell of new heatsink yeah all the oils/grease heats up for the first time and fumes can come out. open the windows for the first 12hrs or so and it should be totally gone after
|
|
|
|